regular testing improves outcomes

The role of testing in education is always going to divide opinion. My position on the matter had always been clear, in that regular low stakes tests or quizzes not only aid learning but also reduce test anxiety by normalising something that, in essence, isn’t normal. Test anxiety is, in part, the result of the unusual nature of the test environment, as well as the fear we all experience when our behaviour, intelligence or knowledge is in some being assessed by someone else. But, like I’ve also pointed out, anxiety isn’t always negative.

Why quizzing works

Testing. Quizzing. Retrieval Practice. Call it what you will, it all boils down to the same thing; recalling from memory something we have learned. When we test ourselves we have to make an effort to recall the information, be it the events that led up to the signing of the Magna Carta, a seminal study in social psychology, the guitar chords for Green Day’s Boulevard of Broken Dreams or parallel parking on a driving test. At some point we are going to have to prove to ourselves or others that we know what to do with the information we’ve learned. This isn’t the opinion of someone taking a side in some imagined educational dichotomy – it’s just the reality of the situation.

The effort of recalling something strengthens the trace of that something and should make it easier to recall it next time. And yes, I’m using should rather than will because there will always be some instances when this doesn’t happen the way we expect it to.

If we gradually increase the length of the gap between recall, the information will eventually stick and from then on it’s going to be relatively permanent – again, I say relatively simply because very few memories are immune from some kind of forgetting or other.  

The meta-analysis

A new meta-analysis seems to support the use of regular quizzing, at least in university students. For the purpose of the study, quizzes were defined as low-stakes assessments of learned material that occurred at least once a week. Marcus Crede and Lukas Sotola of Iowa State University, used data from previously published studies, amounting to 52 classes and nearly 8,000 students. 

The main findings support the effectiveness of regular quizzes, in that students who took them weekly did better on mid-term and end-of-year exams than those who didn’t. They also reduced failure rates in the longer term.

But they also benefitted struggling students the most which is, for me, the most relevant finding.

A few caveats

The best results were seen when students were given immediate feedback, so it’s not just the quiz but also the type of support that is offered afterwards.

It’s also worth noting that multiple choice questions were less effective than constructive response ones (where students had to provide an answer themselves rather than choosing from a list of possibles). That’s not to say that MCQ’s aren’t worth doing, just that they’re not the most effective option.

Another interesting feature of the study is that the researchers didn’t include laboratory type studies, so the results are perhaps more ecologically valid. On the flip side, this also means that possible confounding variables weren’t easy to eliminate.